I just talked to Max in SIMS team, and the conclusion was that we should avoid duplicating metadata in a derived dataset.
It's because there is a risk that a metadata is updated in the parent dataset but not applied in the child dataset.
Instead, Max said they can provide a simple query(through their backend API) that searches all the derived datasets based on their input datasets' metadata.
I thought it was important to you since it'll be the first step to find the input of the analysis software.
I also wrote that down in our repository.
scipp/essreduce#27
And Max also wanted us to write down a use-case scenario so that they can find the best solution.
I'll scribble things here first so that you can review it before I send it to him.
I just talked to Max in SIMS team, and the conclusion was that we should avoid duplicating metadata in a
deriveddataset.It's because there is a risk that a metadata is updated in the
parentdataset but not applied in thechilddataset.Instead, Max said they can provide a simple query(through their backend API) that searches all the
deriveddatasets based on theirinputdatasets' metadata.I thought it was important to you since it'll be the first step to find the input of the analysis software.
I also wrote that down in our repository.
scipp/essreduce#27
And Max also wanted us to write down a use-case scenario so that they can find the best solution.
I'll scribble things here first so that you can review it before I send it to him.